Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
02-18-2015, 01:35 AM   #166
Pentaxian




Join Date: Apr 2007
Location: Sweden
Posts: 1,927
Well put.
Also, if they go the route with four images combined into one with the same resolution but each pixel has full RGB info, then the combined picture takes less memory than the four raw images used to produce it. And combining them would be quick.

02-18-2015, 02:12 AM - 1 Like   #167
Veteran Member




Join Date: Mar 2013
Posts: 320
I think that using pixel shifting technology for resolution enhancements is not a great idea.
If they are thinking to implement that first in a full frame camera, than it is obvious that it will have low resolution (24mp) sensor.
I think it is a much better idea to use sensor shift technology for IQ enhancments (more fine details, less noise) but to stay with the same resolution in a combined image.
I also think that it is not a good idea to combine those images in camera only.
It will be good to have an option like "sensor shift bracketing" and to have ability to post process those images in a computer with a specialized software.
In that case it will be possible to shoot moving subjects and use pixel shifting technology at the same time.

Last edited by banep; 02-18-2015 at 04:32 AM.
02-18-2015, 03:54 AM   #168
Veteran Member
Barry Pearson's Avatar

Join Date: Jul 2013
Location: Stockport
Posts: 870
QuoteOriginally posted by Gimbal Quote
Well put.
Also, if they go the route with four images combined into one with the same resolution but each pixel has full RGB info, then the combined picture takes less memory than the four raw images used to produce it. And combining them would be quick.
I think it is also worth thinking about raw file sizes. The sizes of raw files are dominated by the image data. (Metadata and previews are relatively small when talking about sensors like this).

So (using DNG as an example) the total size would mainly be determined by 3 16-bit integers per pixel, (then losslessly compressed), rather than 4 16-bit integers per pixel, (then losslessly compressed).

(I'm just talking about data sizes, not how the data is organised in the file).
02-18-2015, 06:42 AM - 1 Like   #169
Pentaxian
thibs's Avatar

Join Date: Jun 2007
Location: Belgium
Photos: Albums
Posts: 5,245
QuoteOriginally posted by biz-engineer Quote
If it was about doing super resolution with post processing, it already exists: using longer focal length + stitching.
If super resolution would be provided, it should be computed by the camera and output one high resolution file.
So was the case of HDR but it was integrated into the camera, with a choice Jpeg/Raw and people are happy about it.
I don't care a iota but some will not care about aperture simulator. YMMV.

---------- Post added 18-02-15 at 14:45 ----------

QuoteOriginally posted by banep Quote
I think that using pixel shifting technology for resolution enhancements is not a great idea.
If they are thinking to implement that first in a full frame camera, than it is obvious that it will have low resolution (24mp) sensor.
I think it is a much better idea to use sensor shift technology for IQ enhancments (more fine details, less noise) but to stay with the same resolution in a combined image.
I also think that it is not a good idea to combine those images in camera only.
It will be good to have an option like "sensor shift bracketing" and to have ability to post process those images in a computer with a specialized software.
In that case it will be possible to shoot moving subjects and use pixel shifting technology at the same time.
Resolution of the resulting file and real resolution of an image file are two completely different things.
If you're not convinced, take a look at the film scanners: their output files and real resolution are VERY different.
So yes, you can get higher resolution with a file of the same size (even if Photoshop says the resolution is the same).

02-18-2015, 03:30 PM   #170
Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 27,264
QuoteOriginally posted by thibs Quote
If you're not convinced, take a look at the film scanners: their output files and real resolution are VERY different.
Good point. For scanners the optical system along with negative flatness are the real world limiting factors for resolution. You can over-sample the scan frame all day, but you are not going to make an unsharp image anything more than what it is. My Epson V700 claims 6400 dpi native resolution, but good luck getting better than 2400 in the real world.


Steve
02-19-2015, 11:15 PM   #171
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 9,176
QuoteOriginally posted by starjedi Quote
No new feature at all based on the new interview...
Pentax may not be pursuing traditional super resolution, but based on the below CP+ 2015 presentation, they have apparently implemented the sensel-shift approach that obviates demosaicing:

http://www.youtube.com/watch?feature=player_detailpage&v=0Pa8W-Zer-w#t=1231
02-20-2015, 02:00 AM   #172
Veteran Member




Join Date: Mar 2013
Posts: 320
QuoteOriginally posted by thibs Quote
So was the case of HDR but it was integrated into the camera, with a choice Jpeg/Raw and people are happy about it.
I don't care a iota but some will not care about aperture simulator. YMMV.

---------- Post added 18-02-15 at 14:45 ----------



Resolution of the resulting file and real resolution of an image file are two completely different things.
If you're not convinced, take a look at the film scanners: their output files and real resolution are VERY different.
So yes, you can get higher resolution with a file of the same size (even if Photoshop says the resolution is the same).
Translated to your terminology, when I said "resolution" I thought of a file resolution, not real resolution.
But when I said "better IQ with more fine details within the same resolution", I mean "better real resolution within the same file resolution", ok?.
02-20-2015, 02:42 AM - 1 Like   #173
Veteran Member
Barry Pearson's Avatar

Join Date: Jul 2013
Location: Stockport
Posts: 870
QuoteOriginally posted by Class A Quote
Pentax may not be pursuing traditional super resolution, but based on the below CP+ 2015 presentation, they have apparently implemented the sensel-shift approach that obviates demosaicing:

http://www.youtube.com/watch?feature=player_detailpage&v=0Pa8W-Zer-w#t=1231
It might be useful as an option. But I see various reasons why I would rarely if ever use it. It might treble the sizes of the raw files. It might reduce the frame rate and number of shots in burst mode. And it probably wouldn't work with flash, especially non-Pentax flash such as studio flash.

If it does appear in the camera, I might use it like I use the anti-alias feature of the K-3; in practice, never!

02-20-2015, 06:39 AM   #174
Pentaxian




Join Date: Sep 2010
Posts: 2,646
QuoteOriginally posted by Barry Pearson Quote
It might be useful as an option. But I see various reasons why I would rarely if ever use it. It might treble the sizes of the raw files. It might reduce the frame rate and number of shots in burst mode. And it probably wouldn't work with flash, especially non-Pentax flash such as studio flash.

If it does appear in the camera, I might use it like I use the anti-alias feature of the K-3; in practice, never!
I wouldn't mind triple the raw file size if it works. It's for shots taken on a tripod anyway, not for normal shooting. And not having any moire, regardless of how challenging the shot is, and to have really a great amount of detail without having the camera guess, that would be useful to some. I'd use it when shooting in a situation where I can use it. I think. Hopefully the system will be reasonably fast, so that it can be used outdoors too.
02-20-2015, 11:56 AM   #175
Pentaxian




Join Date: Feb 2015
Photos: Albums
Posts: 3,466
The problem with engineering terms used by marketing staff is that they do not always know what they are talking about. And unfortunately not all customer are in the field of physics , optical engineering, mechanical engineer and electronic engineering.
For instance, the dynamic range was way to much put forward as a key feature of the K-5 when it was out , today we realize that we can see as much difference on the shot as much as it was marketing hype. Then there was the anti-alias simulator thing that what so boldly stated when the k-3 came out. So they are camera features that really make a significant difference when shooting (such as -3ev AF sensors , f/2.8 sensor) and other features are more like gadgets. Now , where does super resolution belong ? I think that super resolution below to the gadget category. We don"t want a camera that takes 8 shots that actually is equivalent to over-sample the same mud of pixels due to the diffraction and all other aberrations for the lenses. What we want is spatial resolution of 4 times 50 lp/mm (ref. nyquist shannon sampling theorem) on a larger sensor so that no optical AA filter is needed to prevent aliasing. Let's say , 24Mpix APS-C (K-3/D7100) or 51MPix FF sensor. For instance you would get much better resolution with a D610 or D750 versus K-3 ; due to the presence of AA filter on D610/D750 and no AA filter on K-3. In order to get the extra resolution for the extra money you'll pay , you'll want to have a FF with 51Mpixels. But I bet that Pentax with first use a FF 24Mpix sensor for the first FF cam , and then introduce a 51Mpix FF later , so that customer still have a good reason to buy (again) one more camera.
02-21-2015, 01:49 AM - 1 Like   #176
Veteran Member
Barry Pearson's Avatar

Join Date: Jul 2013
Location: Stockport
Posts: 870
QuoteOriginally posted by biz-engineer Quote
But I bet that Pentax with first use a FF 24Mpix sensor for the first FF cam , and then introduce a 51Mpix FF later , so that customer still have a good reason to buy (again) one more camera.
I'll bet against 24 MP, partly because it would get bad press, but mainly because it doesn't fit the product range.

They want to sell this camera to the current user base (as well as to new customers). Many people will be using this camera in crop mode with their current lenses at least some of the time. That will reduce the number of pixels used, by up to a factor of 2. 12 MP some of the time would be hard to sell!
02-21-2015, 08:13 PM   #177
Senior Member




Join Date: Nov 2013
Posts: 153
QuoteOriginally posted by Barry Pearson Quote
I'll bet against 24 MP, partly because it would get bad press, but mainly because it doesn't fit the product range.

They want to sell this camera to the current user base (as well as to new customers). Many people will be using this camera in crop mode with their current lenses at least some of the time. That will reduce the number of pixels used, by up to a factor of 2. 12 MP some of the time would be hard to sell!
I agree - I think 36MP is most likely, for the reasons you mention. If the FF is 36MP, then any super-resolution feature is likely to be a gimmick with little real-world use, because the lenses we have to use with it will rarely resolve beyond 36MP except for maybe in a small area in the centre of the frame. Such a feature will only work effectively when shooting static scenes on a tripod - ie. the kind of scene where you want good detail across the frame, so this feature is not going to provide much benefit here either. I suspect that most of us will struggle to resolve to the 36MP native resolution in most shooting scenarios, never mind anything higher.
02-22-2015, 06:35 AM   #178
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,863
QuoteOriginally posted by jonby Quote
I agree - I think 36MP is most likely, for the reasons you mention. If the FF is 36MP, then any super-resolution feature is likely to be a gimmick with little real-world use, because the lenses we have to use with it will rarely resolve beyond 36MP
Being somebody with plenty of experience using 36MP, I beg to disagree. Assuming good practice and good lenses (primes or 2000$ zooms), the sensor is easily outresolved. I can testify color Moiré patterns emerging with fine (almost invisible) fabric all over the frame and with all my lenses. These patterns only emerge when the lens outresolves the sensor. Don't be fooled by MTF50 numbers which only tell a small part of the story. It even happens with Bayer AA filters as all of these are rather weak in modern cameras. Fortunately, said patterns can be treated in software like LR and only occur in a minority of images.

However, aquiring said practice and lenses is a hurdle to make 36+MP a bonus. Not everybody needs that.

Nevertheless, sensel-shift superresolution helps with 36, 50 or even 100MP. Because it gets rid of the demosaicing process and yields better images with no color artefacts.

Especially with architecture and studio product photography, it will soon become a de-facto standard for high resolution dSLRs replacing medium format cameras in this field (except where the tilt feature is mandatory). Maybe less for architecture as pano is an alternative. But I predict the feature even with 100MP full frame cameras.

---------- Post added 22-02-15 at 14:43 ----------

Many seem to believe that 36MP is a genuine choice for the Pentax FF camera.

I disagree. I agree that it is likely it will be 36MP, but not by a large margin.

E.g., Nikon has 2x16, 2x24, 1x36MP in their current lineup, Canon even less below their new1x50MP now. It may be that 50/54MP will be the high end by next year and 36MP replace what is 24MP now. But it will depend on the price point, buffer size and frame rate of the camera Pentax plans to release. And how it positions wrt 645Z.

I expect it to be 36MP too, but I am simply not so sure about it ...
02-22-2015, 07:02 AM   #179
Pentaxian




Join Date: Sep 2010
Posts: 2,646
If our lenses can resolve 24 MP on an APS-C sensor, 36 and more shouldn't be an issue on a FF sensor.
02-22-2015, 01:37 PM   #180
Senior Member




Join Date: Nov 2013
Posts: 153
QuoteOriginally posted by falconeye Quote
Being somebody with plenty of experience using 36MP, I beg to disagree. Assuming good practice and good lenses (primes or 2000$ zooms), the sensor is easily outresolved. I can testify color Moiré patterns emerging with fine (almost invisible) fabric all over the frame and with all my lenses. These patterns only emerge when the lens outresolves the sensor. Don't be fooled by MTF50 numbers which only tell a small part of the story.
Ok I think my statement was a little too sweeping! I'm not going to question your own experiences with your current system and lenses, but my own experience with a range of lenses, including primes and zooms in a range of price brackets and systems, is that not many are capable of resolving well towards the edges of the frame - and that is on sensors with significantly lower resolution than 36MP. But granted, there are lenses available which will do this even at 36MP, and with a static subject, a good tripod and very careful technique, having a super-resolution option could be useful. How many such lenses will be available in K-mount I'm not sure - I can't say that I have a high confidence based on what I have seen from Pentax's DA lenses. I guess my point was that situations where a super-resolution feature will truly offer any real benefit will likely be rare for many photographers.

However, your point about the ability of this technique to reduce demosaicing and moire artifacts (and noise?) is a good one which I missed. This will be of benefit in a wider range of cases. I would certainly find these techniques more useful however, if the output size was still 36MP and not 4x or whatever - is this possible, do you think?

One more thing to mention/clarify: I believe the EM-5 II uses the electronic shutter for super-resolution exposures (please correct me if I'm wrong). This makes sense because even the slightest shutter-induced vibration will knock exposures out of line and reduce the efficiency of the process. Given the relatively large mass of a full frame camera's shutter, would you say that it will be a necessity to employ an entirely electronic shutter for multi-shot exposures in order for this to work reliably? I would guess that it will be.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
architecture, camera, cameras, center, count, ff, file, future pentax, idea, image, images, k3, lenses, mode, models, olympus, pentax, pentax news, pentax rumors, pixel, post, resolution, resolution by pixel, sample, sensor, shift, shutter, technology
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
For Sale - Sold: Pentax KX and ME Super film cameras - both in very good condition HenrikDK Sold Items 6 06-09-2014 05:04 AM
In-Mouth Cameras: The Future of Food Photography? RioRico Photographic Technique 2 10-15-2011 05:07 PM
Pixel pitch and lens resolution surfotog Pentax Medium Format 4 06-09-2011 12:01 PM
Future laid out by Pentax -- no full-frame rparmar Pentax News and Rumors 158 10-23-2008 06:35 PM



All times are GMT -7. The time now is 09:18 PM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top